Registry / ai-ml / catboost

catboost

JSON →
library1.2.10pypypi✓ verified 26d ago

CatBoost is a gradient boosting library developed by Yandex. It currently uses Python package version 0.26.1 corresponding to native library version 1.2.10. Release cadence is irregular, with multiple minor versions per year.

pip install catboost
INSTALL
IMPORT
SIG · CATBOOST
C
catboost
ai-mlpythonv1.2.10
Install
Import
Disk
Pass rate
0/ 10
Env Coverage0 / 10
glibc
3.93.13
musl
3.93.13
Install & Compatibility
Where this runs
tested against v1.2.10 · pip install
no network on importno background threads
Install × environment matrix
Each cell = how many times install + import succeeded across repeated harness runs. Partial = flaky.
glibc = Debian/Ubuntu slim · musl = Alpine Linux
musl
glibc
py 3.10
✕ build_error
1/2 runs
py 3.11
✕ build_error
1/2 runs
py 3.12
✕ build_error
1/2 runs
py 3.13
✕ build_error
1/2 runs
py 3.9
✕ build_error
1/2 runs
Code
Verified usage

Verified import paths — ran on the pinned version, not inferred.

CatBoostRegressor
from catboost import CatBoostRegressor
CatBoostClassifier
from catboost import CatBoostClassifier
Pool
from catboost import Pool

Basic example: train a CatBoostRegressor and predict on sample data.

from catboost import CatBoostRegressor, Pool import numpy as np # Create simple data X = np.random.rand(100, 10) y = np.random.rand(100) train_pool = Pool(X, y) # Train model model = CatBoostRegressor(iterations=10, learning_rate=0.1, verbose=False) model.fit(train_pool) # Predict pred = model.predict(X) print(pred[:5])
catboost --version
Debug
Known issues
gotchaWhen using custom eval metrics on GPU, exceptions from custom metric code may not be propagated correctly (fixed in v1.2.8). Earlier versions might silently ignore errors.
fix
Upgrade to catboost >=0.26.1
affects: 0.26.0 and earlier (native <1.2.8)
gotchaPrediction of type 'Probability' can return NaN on CPUs without SSE4 instruction set (e.g., ARM CPUs). Fixed in node-package v1.27.0 (native library v1.2.10). The Python package 0.26.1 uses native 1.2.10, so it should be fixed, but verify if you see NaN probabilities.
fix
Upgrade to catboost >=0.26.1 or use CPU with SSE4 support
affects: Python packages using native <1.2.10
gotchaOn Windows, MSVC builds did not enable __SSE__ compiler flag, affecting performance of SSE-dependent code paths. Fixed in node-package v1.27.0 (native v1.2.10). Python package 0.26.1 should have this fix.
fix
Upgrade to catboost >=0.26.1
affects: Python packages using native <1.2.10 on Windows
breakingSupport for Python 3.7 dropped in version 1.2.8. Python 3.13 added in same release.
fix
Use Python 3.8+
affects: >0.26.0
gotchaWhen using polars input (added in native v1.2.9), ensure polars is installed separately. CatBoost does not automatically install polars.
fix
pip install polars
affects: 0.26.0 and later
deprecatedjvm applier 'predict' method is deprecated in favor of 'predictTransposed' (added in native v1.2.10).
fix
Use predictTransposed for better performance
affects: Java/CatBoost4j only
Errors
Common errors & fixes
ModuleNotFoundError: No module named 'catboost'
CatBoost is not installed in the current Python environment.
fix
Run: pip install catboost
catboost.CatBoostError: libomp.dylib not found
On macOS, OpenMP library is missing. CatBoost requires libomp to run.
fix
Install OpenMP: brew install libomp
catboost.CatBoostError: GPU support not compiled in: No suitable GPU device found
You installed the CPU-only version of CatBoost (default pip install), which does not include GPU support.
fix
To use GPU, install from source with GPU flags or use a pre-built GPU package (e.g., 'pip install catboost' on systems with CUDA does not automatically enable GPU; use 'pip install catboost' and ensure CUDA toolkit is installed. Alternatively, build from source with -DUSE_GPU=ON). Note: The pip package includes GPU but may fail on some configurations.
Upgrade
Version history
1.2.10latest on PyPI · released Feb 18, 2026
Audit
Dependencies

No dependency data recorded yet.

Agent activity
46 hits · last 30 days
node
40
Amazon
1
OpenAI (training)
1
Resources
catboost — pip install catboost · libregistry